1. Field of the Invention
The present invention relates to a radio frequency (rf) ion source and in particular to a glow discharge source capable of low power operation over a range of pressures, including atmospheric, in air.
2. Discussion of Prior Art
There exists considerable interest in the development of an ion source which is capable of operating under similar conditions to the commercially available electron impact ion source but which is more versatile and more robust than that source. The electron impact ion source is widely used in vapour analysis systems in which it is coupled to a mass spectrometer. In this source ionising particles in the form of electrons are emitted from a heated tungsten wire into a low pressure cavity, which is evacuated to pressures in the region of 10.sup.-4 to 10.sup.-3 Torr. The electrons in this cavity are accelerated by both electric and magnetic fields to an energy where impact of an electron with a sample molecule causes ionisation of that molecule. The electron impact ion source has the disadvantages that it cannot operate at high pressures and that it tends to bum out in oxygen rich environments, making the source unsuitable for use in analysis systems which operate in air at or close to atmospheric pressure.
Additionally, this source has the further disadvantage that it lacks versatility of use since it is effectively limited to the production of positively charged ions in a relatively energetic ionisation process (so called `hard` ionisation) and usually has associated with it sample molecule fragmentation.
There also exists considerable interest in the development of an ion source capable of operating efficiently at atmospheric pressure with air as the discharge gas in which the plasma is maintained and of interfacing with commercially available mass spectrometers. This would allow for the direct sampling of air in order to monitor for the presence of impurity gases, given off for example from some drugs or explosives such as TNT, RDX and PETN.
One known device, which can operate in air at atmospheric pressure, is that described by Zhao and Lubman (Analytical Chemistry Vol 64, No 13, pages 1427-1428 and Vol 65, No 7, pages 866-876) and comprises an insulated tungsten rod driven electrode, of 0.04" diameter and ground at the end to a sharp tip which is the operative end at which a plasma discharge can occur. This electrode is coupled to an rf source and extends into a grounded 1".times.0.8" (diameter) brass cell which forms an effective "plate" electrode. In use the plasma discharge occurs between the operative end of the rod and the cell walls. The sample, ions from which are to be produced and detected, is introduced into the sample-carrying discharge gas as a liquid and carried by the gas into the brass cell where it is ionised. This device however requires a power supply capable of providing the relatively high forward power of approximately 16 Watts (W) to induce the formation and maintenance of a plasma in air at atmospheric pressure. This has the disadvantage that the power supply is relatively costly and bulky.
Furthermore, even at this relatively high forward power this ion source produces only soft (low energy) ionisation and therefore cannot substitute for the electron impact ion source. If hard (high energy) ionisation is needed, then a higher power rf source would be required. This would compound the aforementioned disadvantage since to provide a hard ionising source a power supply which is capable of providing even higher forward powers than those discussed above will be necessary. Moreover, since the plasma generated by the Lubman ion source is stable only over a limited rf range of 125-375 Kilohertz (KHz), then a further disadvantage is that a relatively large ion energy distribution is likely to result which would effectively reduce the resolution of any analysis system incorporating a mass spectrometer. This is because the energy gained from the rf electric field by the ionised particles is, in part, dependent on the frequency of that rf field, as will be readily appreciated by those skilled in the art. If the ionised particles reside in the field long enough to suffer several oscillations of the rf field then their resultant energy will be close to zero, conversely if these particles are formed and ejected from the plasma within the time scale of the rf cycle then their energy will depend on the change in field potential between their formation and ejection. Thus, for a given residence time of an ion created in a radio frequency discharge, the energy distribution of the ejected ionised particles increases as the frequency of the rf field decreases.
Generally in rf ion sources both positive ions and electrons are generated within the plasma. The difference in the mobilities of these charged particles causes a self-bias to develop on the electrode which is capacitatively coupled to the rf power supply. The degree of this self-bias is governed by the geometry of the source and in particular by the relative surface areas of the discharge electrodes, between which a plasma may form. In prior art devices the geometry of the source is such that the surface area of operative end of the driven electrode is small compared with that of the operative end of the grounded (or floating) electrode, which electrode often includes the contacting walls of the ionisation cell. This results in the generation of a negative self-bias. For this reason the driven electrode is customarily termed the "cathode" and the grounded (or floating) electrode the "anode" and therefore throughout this document the terms cathode and anode shall be taken to refer to the driven and grounded (or floating) electrodes respectively.